Oceans of Data Proceedings of the 44th Conference on Computer Applications and Quantitative Methods in Archaeology

Edited by Mieko Matsumoto and Espen Uleberg Archaeopress Publishing Ltd Summertown Pavilion 18-24 Middle Way Summertown Oxford OX2 7LG www.archaeopress.com

ISBN 978 1 78491 730 2 ISBN 978 1 78491 731 9 (e-Pdf)

© Archaeopress and the authors 2018

Language Editing: Rebecca Cannell Cover Design: Bjarte Einar Aarseth

All rights reserved. No part of this book may be reproduced, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners.

This book is available direct from Archaeopress or from our website www.archaeopress.com Contents

Foreword...... v

Introduction...... 1 Oceans of Data: Creating a Safe Haven for Information...... 3 Christian-Emil ORE Theorising the Digital: A Call to Action for the Archaeological Community...... 11 Sara PERRY and James Stuart TAYLOR

Ontologies and Standards...... 23 Is that a Good Concept?...... 25 George BRUSEKER, Maria DASKALAKI, Martin DOERR, and Stephen STEAD Sculptures in the Semantic Web Using Semantic Technologies for the Deep Integration of Research Items in ARIADNE...... 33 Philipp GERTH, Dennis Mario BECK, Wolfgang SCHMIDLE, and Sebastian CUY Formalization and Reuse of Methodological Knowledge on Archaeology across European Organizations... 45 Cesar GONZALEZ-PEREZ, Patricia MARTÍN-RODILLA, and Elena Viorica EPURE Linked Open Data for Numismatic Library, Archive and Museum Integration...... 55 Ethan GRUBER Sustainability = Separation: Keeping Database Structure, Domain Structure and Interface Separate...... 63 Ian JOHNSON Systematic Literature Review on Automated Monument Detection: A Remote Investigation on Patterns within the Field of Automated Monument Detection...... 69 Karl Hjalte Maack RAUN and Duncan PATERSON Bioarchaeology Module Loading…Please Hold. Recording Human Bioarchaeological Data from Portuguese Archaeological Field Reports...... 85 Ana Lema SEABRA, Filipa Mascarenhas NETO, and Cristina BARROSO-CRUZ Methodological Tips for Mappings to CIDOC CRM...... 89 Maria THEODORIDOU, George BRUSEKER, and Martin DOERR An Ontology for a Numismatic Island with Bridges to Others...... 103 Karsten TOLLE, David WIGG-WOLF, and Ethan GRUBER Integrating Analytical with Digital Data in Archaeology: Towards a Multidisciplinary Ontological Solution. The Salamis Terracotta Statues Case‑Study...... 109 Valentina VASSALLO, Giusi SORRENTINO, Svetlana GASANOVA, and Sorin HERMON

Field and Laboratory Data Recording and Analysis...... 119 Integrated Methodologies for Knowledge and Valorisation of the Roman Casinum City...... 121 Michela CIGOLA, Arturo GALLOZZI, Leonardo PARIS, and Emanuela CHIAVONI A Multidisciplinary Project for the Study of Historical Landscapes: New Archaeological and Physicochemical Data from the ‘Colline Metallifere’ District...... 135 Luisa DALLAI, Alessandro DONATI, and Vanessa VOLPI From Survey, to 3D Modelling, to 3D Printing: Bramante’s Nymphaeum Colonna at Genazzano...... 147 Tommaso EMPLER and Adriana CALDARONE Towards a National Infrastructure for Semi‑Automatic Mapping of Cultural Heritage in Norway...... 159 Martin KERMIT, Jarle Hamar REKSTEN, and Øivind Due TRIER

i Experiments in the Automatic Detection of Archaeological Features in Remotely Sensed Data from Great Plains Villages, USA ...... 173 Kenneth L. KVAMME Interpolating 3D Stratigraphy from Indirect Information...... 185 Lutz SCHUBERT, Ana PREDOI, and Keith JEFFERY Closing a Gap with a Simple Toy: How the Use of the Tablet Affected the Documentation Workflow during the Excavations of the Rozprza Ring–Fort (Central Poland)...... 197 Jerzy SIKORA and Piotr KITTEL Supercomputing at the Trench Edge: Expediting Image Based 3D Recoding...... 207 David STOTT, Matteo PILATI, Carsten MEINERTZ RISAGER, and Jens-Bjørn Riis ANDRESEN Semi‑Automatic Mapping of Charcoal Kilns from Airborne Laser Scanning Data Using Deep Learning...... 219 Øivind Due TRIER, Arnt-Børre SALBERG, and Lars Holger PILØ Documenting Facades of Etruscan Rock‑Cut Tombs: from 3D Recording to Archaeological Analysis...... 233 Tatiana VOTROUBEKOVÁ

Archaeological Information Systems...... 243 Fasti Online: Excavation, Conservation and Surveys. Twelve Years of Open Access Archaeological Data Online...... 245 Michael JOHNSON, Florence LAINO, Stuart EVE, and Elizabeth FENTRESS DOHA — Doha Online Historical Atlas...... 253 Michal MICHALSKI, Robert CARTER, Daniel EDDISFORD, Richard FLETCHER, and Colleen MORGAN Digital Archives — More Than Just a Skeuomorph...... 261 Emily NIMMO and Peter MCKEAGUE When Data Meets the Enterprise: How Flanders Heritage Agency Turned a Merger of Organisations into a Confluence of Information...... 273 Koen VAN DAELE, Maarten VERMEYEN, Sophie MORTIER, and Leen MEGANCK

GIS and Spatial Analysis...... 285 Crossroads: LCP — Model Testing and Historical Paths During the Iron Age in the North–East Iberian Peninsula (4th to 1st Centuries BC)...... 287 Joan Canela GRÀCIA and Núria Otero HERRAIZ Boundaries of Agrarian Production in the Bergisches Land in 1715 AD...... 299 Irmela HERZOG Geometric Graphs to Study Ceramic Decoration...... 311 Thomas HUET Vertical Aspects of Stone Age Distribution in South–East Norway...... 325 Mieko MATSUMOTO and Espen ULEBERG

3D and Visualisation...... 337 Emerging Technologies for Archaeological Heritage: Knowledge, Digital Documentation, and Communication...... 339 Martina ATTENNI, Carlo BIANCHINI, and Alfonso IPPOLITO New Actualities for Mediterranean Ancient Theaters: the ATHENA Project Lesson...... 353 Carlo BIANCHINI, Carlo INGLESE, and Alfonso IPPOLITO Archaeology and Augmented Reality. Visualizing Stone Age Sea Level on Location...... 367 Birgitte BJØRKLI, Šarūnas LEDAS, Gunnar LIESTØL, Tomas STENARSON, and Espen ULEBERG A Virtual Reconstruction of the Sun Temple of Niuserra: from Scans to ABIM...... 377 Angela BOSCO, Andrea D’ANDREA, Massimiliano NUZZOLO, Rosanna PIRELLI, and Patrizia ZANFAGNA

ii A 3D Digital Approach for the Study and Presentation of the Bisarcio Site...... 389 Paola DERUDAS, Maria Carla SGARELLA, and Marco CALLIERI The Role of Representation in Archaeological Architecture...... 399 Mario DOCCI, Carlo INGLESE, and Alfonso IPPOLITO Digital Archaeological Dissemination: Eleniana Domus in Rome...... 409 Tommaso EMPLER On Roof Construction and Wall Strength: Non-Linear Structural Integrity Analysis of the Early Bronze Age Helike Corridor House...... 421 Mariza Christina KORMANN, Stella KATSAROU, Dora KATSONOPOULOU, and Gary LOCK An Exploratory Use of 3D for Investigating a Prehistoric Stratigraphic Sequence...... 433 Giacomo LANDESCHI, Jan APEL, Stefan LINDGREN, and Nicolò DELL’UNTO Les gestes retrouves: a 3D Visualization Approach to the Functional Study of Early Upper Palaeolithic Ground Stones...... 447 Laura LONGO, Natalia SKAKUN, Giusi SORRENTINO, Valentina VASSALLO, Dante ABATE, Vera TEREHINA, Andrei SINITSYN, Gennady KHLOPACHEV, and Sorin HERMON Enhancing Archaeological Interpretation with Volume Calculations. An Integrated Method of 3D Recording and Modeling...... 457 Giulio POGGI and Mirko BUONO 3D Spatial Analysis: the Road Ahead...... 471 Martijn VAN LEUSEN and Gary NOBLES

Complex Systems Simulation...... 479 Weaving the Common Threads of Simulation and Formation Studies in Archaeology...... 481 Benjamin DAVIES Evolving Hominins in HomininSpace: Genetic Algorithms and the Search for the ‘Perfect’ Neanderthal..495 Fulco SCHERJON An Agent‑Based Approach to Weighted Decision Making in the Spatially and Temporally Variable South African Paleoscape...... 507 Colin D. WREN, Chloe ATWATER, Kim HILL, Marco A. JANSSEN, Jan C. DE VYNCK, and Curtis W. MAREAN

Teaching Archaeology in the Digital Age...... 523 Archaeological Education for a Digital World: Case Studies from the Contemporary and Historical US.....525 Anna S. AGBE-DAVIES Teaching Archaeology or Teaching Digital Archaeology: Do We Have to Choose?...... 533 Sylvain BADEY and Anne MOREAU DOMUS: Cyber‑Archaeology and Education...... 541 Alex DA SILVA MARTIRE and Tatiana BINA Digital Data Recording at Circus Maximus: A Recent Experience...... 547 Alessandro VECCHIONE and Domenica DININNO Teaching GIS in Archaeology: What Students Focus On...... 555 Mar ZAMORA MERCHÁN and Javier BAENA PREYSLER

iii iv Mieko Matsumoto and Espen Uleberg (eds) 2018. CAA2016: Oceans of Data Proceedings of the 44th Conference on Computer Applications and Quantitative Methods in Archaeology. Oxford: Archaeopress

A Multidisciplinary Project for the Study of Historical Landscapes: New Archaeological and Physicochemical Data from the ‘Colline Metallifere’ District

Luisa DALLAI1, Alessandro DONATI2, and Vanessa VOLPI2 1Department of Historical Sciences and Cultural Heritage, University of Siena, 2Department of Biotechnology, Chemistry and Pharmacy, University of Siena, Italy Corresponding author: [email protected]

Abstract In this paper we will present some results obtained by a combined multidisciplinary approach to the study of a territory located in the southern part of , the ‘Colline Metallifere’ district. The area has been studied from different points of view, and there are many written contributions concerning the evolution of its human landscape over the centuries. In this paper we are focusing particularly on the use of chemical analyses on crucial historical areas. The considerable amount of information re- trieved and the unquestionable value of the data obtained by portable X-ray fluorescence (pXRF) analyses are now opening new fields to the use of the portable instrument. The new opportunities offered by the ‘NeuMed’ ERC project, based in the University of Siena, will allow the testing of the reliability of this technique on a wider scale, in multidisciplinary research that aims to reconstruct the historical landscapes of the area. Keywords: interdisciplinary methodology, archaeological surveys, physical-chemical analysis, pXRF, mining territories

The Colline Metallifere district: a multidisciplinary the central centuries of the medieval period, developed open air laboratory (L.D.) their control over the underground resources through fortified settlements placed close to the mining fields. In recent years, the Colline Metallifere district The archaeological research undertaken since the (southern Tuscany) has become a multidisciplinary 1980s and the excavation of ‘mining castles’, such as the study area in which a combined archaeological, well-known castle of Rocca San Silvestro or the castle geological, and environmental approach is providing of Rocchette Pannocchieschi, have demonstrated that new and interesting evidence for the reconstruction of the development of local lordships in this geographical historical landscapes. The territory is very well known context was deeply linked to the control and from the numerous research projects undertaken over exploitation of metals that could be used for coinage the years on archaeological and historical targets and (Bianchi, 2010). mining deposits (Dallai and Francovich, 2005). The hills which give the area its name cover part of Livorno and Since then, a large number of research projects have Grosseto provinces; the territory includes the coast and been carried out with the aim of studying the possible the immediate hinterland of the Gulf of Follonica, an relations between settlements and mineral resources, important working hub for the iron oxides (hematite) with particular reference to settlement patterns and that have been exploited on Elba Island since the the control over the cycle of production through the Etruscan period. centuries. Extensive surveys on territorial samples have been carried out in order to define the main Ever since that period, the entire area has been aspects of mining exploitation and metal production; historically very important for mining activities. During the data collected has been processed with the aim the medieval period in particular, the exploitation was of defining a broad picture of the territory, both from especially focussed upon mixed sulphides deposits of an archaeological, geological and environmental copper, lead and silver; the consistent ore bodies were perspective (Benvenuti et al., 2014). worked in order to obtain the metals needed for coin production. Out of this wide territory, specific targets have been selected for a multidisciplinary research; these ‘pilot’ These rich deposits and their exploitation attracted the studies started with in‑depth research carried out interest of important aristocratic families, which, since on mining fields (particularly in the mining areas of

135 M. Matsumoto and E. Uleberg (eds). CAA2016: Oceans of Data

Figure 1. Location of sites mentioned in the text.

Serrabottini and Niccioleta, close to , as on other key samples that have been discussed in Grosseto province). In order to get the maximum previous papers (Dallai et al., 2015b), a multidisciplinary amount of useful information, a combination of approach has been first planned and then undertaken underground surveys, fieldwork and XRF analyses on directly on site. Finally, the obtained results have been water-borne sediments and soils have been planned. The examined and analysed in order to obtain increased data collected has clarified the technical characteristics information. The final goal of our project is, in fact, of the production and the environmental impact to provide an historical and environmental landscape of ancient mines on the surrounding area. Finally, reconstruction based on multidisciplinary data sets; chemical analyses carried out on waste material has merging historical data with chemical-physical results revealed a skilful selection operated at the mine head in we are gradually adding crucial pieces of information order to eliminate useless minerals, zinc in particular to the picture of one of the most important mining (Aranguren et al., 2007; Dallai et al., 2015a; Dallai et district of the Mediterranean area. al., 2013). A combination of underground surveys, fieldwork and XRF analyses on fluvial sediments and From the different case studies, by combining soils explored the technical aspects of the production archaeological and physico-chemical data (pXRF in and the environmental impact of ancient mines in the particular) we are gradually building up a solid database surrounding area (Aranguren et al., 2007). that is helping to define the historical outlines of the Colline Metallifere landscape. Moreover, building upon The multidisciplinary approach has focused not only previous research experience, pXRF analyses are now on territorial samples, but has also included the study playing an essential role in the research strategies of of specific archaeological sites; some of which are the 5 year ERC project ‘NeuMed Origins of a new economic medieval settlements located in the core of mining union, 7th–12th centuries: resources, landscapes and political areas (that is the case of the church of San Niccolò, near strategies in a Mediterranean region’ based in the University ); some are production sites (as for the alum of Siena (P.I. Prof. Richard Hodges). The reconstruction production site of Monteleo, close to Monterotondo of the historical features in the landscape becomes Marittimo); whilst others are sites based in the coastal crucial to understand the deep changes that occurred plan, near the shore, that have been identified by in settlement patterns, trade routes and economical archaeological surveys and remote sensing techniques background of the area between Late Antiquity and the and recently excavated (i.e. the Carlappiano site, near 12th century. Piombino) (Figure 1). On all these specific places, as well

136 L. DALLAI et al: A Multidisciplinary Project for the Study of Historical Landscapes

Investigation methods (V.V., A.D.)

Physico-chemical analysis of different environmental matrices is becoming a fundamental tool for multi-scale archaeological prospections. From the enrichment or depletion of certain elements or molecules in soil, stream sediments, groundwater etc., their spatial distribution and their statistical treatment gives the chemical fingerprint of a territory with possible genuine correlations with ancient human activities (Oonk et al., 2009; Anguilano et al., 2010).

Recently, pXRF has been used for the quantification of major and minor elements in soils, stream sediments and artefacts, and it emerged as the elective technique for in-situ analysis. In fact, the pXRF instrument can be directly used on the surface of untouched environmental matrices and immobile artefacts. It can be also used for laboratory analysis with treated Figure 2. pXRF laboratory station. (dried, milled, sieved) samples. The major features of pXRF analysis are the non-destructive nature of analysis, the speed of operations, the capability of on- On the other hand, the ancient mining field can also site measurement, and the immediate availability of be investigated spatially and directly. In this case, analytical results (Potts and West, 2008; Shugar and the concentration of elements in geo-referenced soil Mass, 2012). The protocol can be used in multi‑scale samples gives the chemical fingerprint of the area. investigations (in-situ and medium‑large territorial scale) with both predictive and descriptive goals using Both laboratory and on-site pXRF analysis were slightly different techniques. performed using an Olympus DELTA-premium handheld pXRF analyser, equipped with a 40kV tube, a In-situ studies were conducted collecting data using a ‘large area SDD detector, accelerometer and barometer 2 pre-determined grid frame of 1 m (1 m × 1 m) areal for atmosphere pressure corrections for light elements elements. Inside each areal element the concentration measurements. The laboratory measurements were of a chemical species was given by the average done with the instrument mounted on a fixed station of three single pXRF measurements (Dallai et al., (Figure 2). To ensure the quality of the data obtained, 2015a). Depending on the required sampling density, operating protocols was used (EPA, 2007). All laboratory the number of pXRF sample points in one of areal samples were collected in the same spot as the on-site element could be much higher, often following the measurements; they were dried at room temperature, concurrent excavation of single units. In this context sieved at 125μm and placed in the appropriate sample the measurements done directly on archaeological holders. Data were acquired with the ‘Soil Mode’ finds or archaeological structures (architectural (3–beam) of the instrument which utilises Compton elements, furnaces, walls, mortars, slags etc.) can also Normalization for low concentrations (PPM to 3%) of be included, which is a great help for their functional elements in light matrices. The result for each sample characterisation. was the average of the three measurements. Chemical data was finally geo-referenced together with the In the medium‑large territorial scale the archaeological information in a GIS application (QGIS) sampling density is lower and usually follows the used to produce distribution maps. geomorphological elements of the territory. The fluvial stream sediments were proven to be the most Medium-large territorial scale (V.V., A.D.) useful environmental matrix for studying the chemical anomalies due to ancient industrial settlements. This In this work we present the results obtained in two sites because the streams and rivers are the collector of that have recently been investigated. One of them was the contaminated material originated by one or more the ancient Serrabottini mining area, and the other is different sources within their own drainage basin. In the Monte Gai site; both of them are close to the medieval particular, the production activities exploited in this town of Massa Marittima. Here, the major activity mining area can be traced in detail from the extraction was the extraction and processing of mixed sulphide of the mineral ores to the smelting and refining of the ores for the production of copper. The chemical data metals. collected in mining dumps areas confirmed the major

137 M. Matsumoto and E. Uleberg (eds). CAA2016: Oceans of Data

Figure 3. Serrabottini (Massa Marittima, GR). The chemical data collected through pXRF analyses on ancient and modern mining dumps have revealed the main goal of the ancient mining exploitation (particularly Cu), as well as the contamination of the soil.

goal of the exploitation (copper production) as well as It was noteworthy that by observing geo-referenced the extent of heavy metal pollution and its diffusion. data on the site map (Figure 4B), a spatial separation In the Figure 3, the concentration of Cu, Zn and Pb are of the clusters emerges. In fact, Cluster‑1, with higher reported on the map of the site, showing the position concentration of Cu, Zn and Fe was mainly restricted of the large mine tails dumps that are located in a in two small areas with respect to the Cluster‑2 data. strategic position with respect to the extraction sinks This fact could be explained with the hypothesis that in (red circles). The data also revealed that the diffusion these two areas a different production activity (namely of the contamination is quite restricted to the sites and roasting of mineral ores) was accomplished. Cluster‑3 its spreading is poorly enhanced by the weathering of was constituted by only two samples with very high mining residues. lead concentrations.

Regarding the Serrabottini site, a k-mean clustering This result can also be explained by considering that analysis was also performed. In Figure 4, ‘A’ reports during the first phase of the metallurgical process for one of the ten different correlation diagrams obtained copper production, the roasting of mixed sulphide ores, by the k-mean clustering, in which different colours induced a large loss of the most volatile metals: arsenic, represented samples that were grouped by chemical antimony and lead; while zinc was not completely lost similarity. in this phase. To eliminate almost all impurities and zinc, after the roasting, a preliminary phase of smelting

138 L. DALLAI et al: A Multidisciplinary Project for the Study of Historical Landscapes

Figure 4. Statistical treatment of the data (K-means clustering analysis). In Figure A the correlation diagram of Cu-Zn; in Figure B, the georeferenced data on the site map. produced a copper matte that was roasted a second ‘Canonica di San Niccolò’ (Montieri — GR) (L.D., A.D.) time before the final smelting. Montieri district has been one of the most prominent In-situ analyses and results (L.D., A.D., V.V.) silver, lead and copper mining areas of the Colline Metallifere. It was certainly at the heart of a complex Regarding in-situ investigations, here we present three system of ore-working and mineral production, key sites showing relevant data, with interesting probably connected to the activities of the neighbouring connections to the archaeological evidences: the castles. Its’ importance is documented by the presence ‘Canonica di San Niccolò’ site (Montieri — GR); the of a mint which, between the end of the 12th century ‘Allumiere di Monteleo’ site ( and the first half of the following century, struck coins — GR) and the Carlappiano site (Piombino — LI). on behalf of the Volterra bishop.

139 M. Matsumoto and E. Uleberg (eds). CAA2016: Oceans of Data

Figure 5. The intra-situ pXRF analyses at the Canonica site; the sampling grid.

NW from the village, on the slopes of the so called ‘Poggio’, a systematic set of analyses have been carried out on the archaeological site of the Canonica di San Niccolò (a parish church), a peculiar church building with six apses, with adjacent spaces and buildings dedicated to different activities. The documents attest the existence of the site from at least 1133, while the dig provides evidence of an initial occupation of the terraced site probably in the period between the 9th and 10th centuries. The final abandonment of the site Table1. Comparison of simple statistic parameters for Fe, Cu, occurred just before the 15th century (Benvenuti et al., Sn, Pb data obtained in the Canonica area (Montieri — GR). 2014).

At the Canonica site, a very large sampling grid frame of 1 m2 was applied, covering a large part of the AREA 3000 had principally an agricultural function excavation area (AREA 3000 and AREA 2000) (Figure whilst AREA 2000 was a production area where different 5). The chemical elements for this type of analysis types of craft activities were conducted. In this area considered as significant ‘tracer’ elements, useful to the excavation campaigns had already identified the identify archaeometallurgical production, are Pb, Fe, remnants of forging activities and other production Cu and Sn. evidence.

Simple statistic parameters for Fe, Cu, Sn, Pb data A more accurate analysis was obtained by observing regarding the Canonica site showed lower values for the spatial distribution of the metal concentrations AREA 3000 with respect to AREA 2000, both for averaged (Figure 6). In the western part of AREA 2000, the higher and peak values (Table 1). This fact indicates that concentrations of Pb and Fe confirmed the presence of different activities were carried out in the two parts. a forge, while in the centre, high values of Sn and Cu

140 L. DALLAI et al: A Multidisciplinary Project for the Study of Historical Landscapes

Figure 6. Results of intra-situ pXRF analyses. The ring circumference is proportional to the concentration of the considered chemical element. Analyses have highlighted high concentrations of Sn (A) and Cu (B) on the centre of the area 2000.

Figure 7. pXRF analyses have been used for both descriptive and predictive goals. In Figure A the intra-situ analyses before the discovery of the bell‑kiln remains (Figure B) (credits: LTTM, University of Siena).

identified the presence of an activity linked with the the small medieval village of Montieri, one of the most use of an alloy of these two metals. The subsequent dig important silver extraction sites during the Middle campaign discovered the remains of a bell-kiln located Ages. Previous investigation took into account the metal exactly in the area of the pXRF anomaly (Figure 7, concentrations in the stream and river sediments of the Dallai et al., 2015a). No evidence was found regarding area. Using the approach on medium-large territorial the anomalous presence of Pb related to the silver scales, a wide range of information about ancient mining, production. roasting and smelting activities was obtained for the area around the village (Dallai et al., 2015a). Contrary to In this case, the in-situ analysis has confirmed and previous work, the rivers around the Canonica did not demonstrated the reliability of this technique and present evidence of metal contamination. In summary, the descriptive and predictive power for the study of the combination of physico-chemical and historical archaeological sites with a production ‘vocation’. analysis have been definitely useful to understand the nature of the metallurgical activities undertaken near The reported study performed at the Canonica of San the village of Montieri, as well as to detail the nature of Niccolò, was a part of a wider project started few years working activities that took place at the Canonica site. ago in order to understand the productive history of In this case, the absence of metal processing evidence

141 M. Matsumoto and E. Uleberg (eds). CAA2016: Oceans of Data

is providing important new information on adopted technologies. Most of the structures which have been brought to light may be ascribed to the so-called ‘Renaissance phase’ of the site, i.e. to between the end of the 15th century and the first half of the 16th century, a period of documented renewal of alum mining and processing in numerous areas of Tuscany (Boisseuil and Chareille, 2009) (Figure 8).

Despite the written sources, the archaeological investigation revealed a longer history on the site, with evidence of production activities starting much earlier than the 15th century. The remnants of wall structures Figure 8. The battery of the early‑modern alum furnaces in irregular limestone blocks bonded with soil, most at the Monteleo site. The red ring indicates the area where likely relating to a furnace, refer to this earlier period pXRF analyses were conducted. and were obliterated by the 16th century masonry of the alum works. Radiocarbon dating performed on different charcoal samples have proven that the furnaces occupying the terrace, as well as other excluded its direct involvement in the silver economy, productive structures and forges identified on the site, which was the most particular feature of the area were used in the late 13th century. during the medieval centuries. In-situ pXRF analysis (i.e. analysis planned inside a ‘Allumiere di Monteleo’ (Monterotondo Marittimo — GR) single archaeological site), were conducted during the (L.D.) excavation campaigns in order to understand the kind of production related to these ancient remains. Whilst The Monteleo alum works are located in the NW on the Canonica site the pXRF analysis was conducted reaches of the Monterotondo Marittimo municipality principally on the soil, at Monteleo this technique was and extend along the two banks of a local stream, the also applied to the remains of the productive structures, waters of which played an essential role in numerous as well as on finds (slags and metal droplets). The pXRF steps of the production process (Dallai, 2014). From analysis, in particular has identified former metal an archaeological point of view, the discovery of this production involving copper and silver sulphides. site and of a non-metallic working cycle in the Colline Data analysis showed the presence of high anomalies Metallifere district dating to the late medieval period, of Cu (2000 mg/Kg) and Fe (3000 mg/Kg) in specific such as the alunite one, has enlarged upon the already and restricted areas of the site, linked respectively to a detailed evidence of pre-industrial extraction activities copper furnace and a forge (Figure 9). The pXRF analysis in the area, which until now referred almost exclusively of several metallic fragments and residues from several to the cultivation of mixed sulphides. stratigraphic units show high concentrations of S, Ca, Cu, Fe, Pb, Sn, Zn and Ag; copper is always present with The raw material worked in this site is alunite, an high concentrations (60 wt%) (Dallai and Volpi, 2015). aluminium potassium sulphate (Kal3(SO4)2(OH)6) that is practically insoluble in the natural state. Nevertheless, Carlappiano (Piombino — LI) (V.V.) it is possible to transform alunite into high quality alum through a process of roasting, maceration, leaching While the previous examples proposed have shown the and crystallisation known since medieval times and high potential of analyses performed on archaeological described in detail in two renown 16th century technical excavation and territorial samples, new research treatises: De la Pirotechnia by the Sienese, Vannoccio perspective are now open to a multidisciplinary Biringuccio published in 1540, and De re metallica by approach within the frame of the ERC Advanced Giorgius Agricola, published in 1556: to date, detailed European Research Project Neumed, mentioned above. reconstructions of the technical procedures and The project, started in October 2015, has selected the tentative functional interpretations of known material first targets on which archaeological excavation as well finds have mostly been based on these descriptions. as environmental analysis have been planned and the Carlappiano site is one of these. Prior to the dig, remote There are currently only a few but significant case sensing (i.e. magnetometer, drone and historical aerial studies on archaeological sites linked to the production photographs analyses) as well as pXRF measurements of alum from alunite, and Monteleo is one of these. The have been carried out, providing promising initial excavation, begun in 2008 by the University of Siena, results. The site is located on a dune close to the shore

142 L. DALLAI et al: A Multidisciplinary Project for the Study of Historical Landscapes

Figure 9. At Monteleo (Monterotondo Marittimo, GR), a late medieval alum productive site, pXRF analyses and archaeological data have identified a former metal production involving copper and silver sulphides. In Figures A and B, the archaeological remains of the late 13th century productive phase. In Figure D, the pXRF analyses indicates the presence of high concentrations of copper (credits: LTTM, University of Siena). line, on the boundary of what once was a coastal lagoon, coverage, 1938–2015). For this investigation, a particular not far from the city of Piombino. From historical aerial sampling strategy was developed. Six transects, with photos, a dry area of about 8 ha surrounded by a dark North — South direction were investigated, having one almost circular sign was clearly visible. The analysis measurement every 20 meters; some measurements of 19th century historical cartography allowed us to were taken also in the area surrounding the site. The recognise a hydrographic system composed of two chemical elements analysed were As, Cu, Zn, Fe, Ca different streams, which met exactly at the height of and Pb (Figure 10). It is noteworthy that the element the dune, very close to the site, and lead to the sea with distribution showed the presence of two different soil a seemingly large mouth. compositions, and that the separation line between them corresponded almost perfectly to the aerial The pottery and finds collected on the surface offered anomaly boundary. A hypothesis is these differences a wide chronological range: some were dated to the could be related to the function of the settlement Bronze Age, many others could be referred to the that the archaeological excavation has linked to salt Roman periods (imperial and late antique), and to production (Dallai et al., in prep). the early medieval (ca 9th AD) and central medieval centuries (13th–14th) (Dallai et al., 2003; Marasco, 2013). Conclusions (L.D., A.D., V.V.)

Analyses using pXRF were performed on the The use of pXRF on archaeological contexts is gradually Carlappiano area, by using both on-site and laboratory gaining credit, given to the amount of case studies that methods. The on-site analyses were conducted both are providing new evidence for the reliability of this inside and outside the anomaly previously observed technique. The multidisciplinary approach developed by aerial images (interpretation of vertical historical in the Colline Metallifere project and the data collected,

143 M. Matsumoto and E. Uleberg (eds). CAA2016: Oceans of Data

Figure 10. On the Carlappiano site (Piombino — LI), the archaeological dig has been anticipated by a multidisciplinary strategic approach, in order to map archaeological remains and relic landforms. These were visible both on‑site (air photo interpretation of vertical historical coverage, 1938–2015) and in the surrounding area. In particular pXRF has highlighted how the roundish shape corresponds to an effectively different soil composition. In Figures B‑E: Fe, Ca, Cu and Zn concentrations and in the legend are the concentration range considered for each element.

144 L. DALLAI et al: A Multidisciplinary Project for the Study of Historical Landscapes that we have briefly presented, demonstrate that Russo, L., and Vanessa, V. (in prep) ‘Investigations pXRF can greatly help in understanding the nature of at Carlappiano: new archaeological findings, and activities carried out on ancient sites, as well as defining man-made and natural landscapes’, in Hodges, R. the use of productive structures and entire territories. and Bianchi, G. (eds) Origins of a new economic union Moreover, the proper use of pXRF techniques and the (7th–12th centuries): resources, landscapes and political careful evaluation of the analytical results based on strategies in a Mediterranean region. Il progetto nEU- the co-operation of different research expertise, can Med ed i risultati delle prime ricerche 2015–2017. not only describe the environmental and historical Dallai, L., Bianchi, G., Donati, A., Trotta, M., and Volpi, evidences, but can be used as a powerful predictive tool V. (2015a) ʹLe analisi fisico-chimiche territoriali which can drive the research. ed “intra-sito” nelle Colline Metallifere: aspetti descrittivi, “predittivi” e prima interpretazione Through the combination of different data sets, including storica dei dati ʹ, in Arthur, P. and Imperiale, M. L. those that can be obtained from pXRF analyses, the (eds) VII congresso nazionale di archeologia Medievale reconstruction of historical landscape can definitely (Vol. 1). Firenze: All’Insegna del Giglio, pp. 389–394. be more detailed; the amount of measurements that Dallai, L., Donati, A., Bardi, A., and Fanciulletti, S. (2013) can be performed by the user-friendly pXRF has to be ʹArcheologia e chimica per il patrimonio minerario considered as an advantage on wider scale projects, (Ar.Chi.Min.). Un nuovo approccio multidisciplinare like the one we have described. From the consistent allo studio dei contesti archeominerari del experience gained with the multidisciplinary study of comprensorio massetanoʹ, in Galeotti, G. and the mining territories of the Colline Metallifere we can Paperini, M. (eds), Città e Territorio. Conoscenza, tutela now experiment with the potential of pXRF on different e valorizzazione dei paesaggi culturali. Livorno: Debatte kinds of sites, such as the coastal ones selected by the editore, pp. 86–91. Series: Collana Confronti, Vol. 1. NeuMed ERC project, along with Carlappiano, which is Dallai, L., Donati, A., Volpi, V., and Bardi, A. (2015b), provided the first, promising case study. ʹArchaeological and physicochemical approaches to the territory: on-site analysis and multidisciplinary References databases for the reconstruction of historical landscapesʹ, in Stefano, C., Scopigno, R., Carpentiero, Anguilano, L., Timberlake, S., and Rehrn, T. (2010) G., and Cirillo, M. (eds) CAA2015. Keep the Revolution ʹAn early medieval lead-smelting bole from Banc Going >>> Proceedings of the 43rd Annual Conference Tynddol, Cwmystwyth, Ceredigionʹ, Historical on Computer Applications and Quantitative Methods in Metallurgy, 44 (2), pp. 85–103. Archaeology (Vol. 1). Oxford: Archaeopress, pp. 177– Aranguren, B., Bagnoli, P., Dallai, L., Farinelli R., and 185. Negri, M. (2007) ʹSerrabottini (Massa Marittima, Dallai, L. and Francovich, R. (2005) ʹArcheologia di GR): indagini Archeologiche su un antico campo miniera ed insediamenti minerari delle Colline minerarioʹ, Archeologia Medievale, XXXIV, pp. 79–94. Metallifere grossetane nel Medioevoʹ, in Ciardi, M. Benvenuti, M., Bianchi. G., Bruttini, J., Buonincontri, and Cataldi, R. (eds) Il calore della terra. Contributo alla L., Chiarantini, L., Dallai, L., Di Pasquale, G., Donati, storia della geotermia in Italia. Pisa: ETS, pp. 126–142. A., Grassi, F., and Pescini, V. (2014) ʹStudyng the Dallai, L., Giannoni, L., Patera, A., Shepherd, E.J., and Colline Metallifere mining area in Tuscany: an Zanini, E. (2003) ʹ Il Vignale ritrovatoʹ, in Mascione, interdisciplinary approachʹ, in Silvertant, J. (ed.) C. and Patera, A. (eds) Materiali per , 2. Research and preservation of ancient mining areas. 9th Firenze: All’Insegna del Giglio, pp. 281–310. International Symposium on archaeological Mining Dallai, L. and Volpi, V. (2015) ʹRisorse del sottosuolo e History, Valkenburg: Silvertant Erfgoedprojecten, cicli produttivi: allume, rame e argento. Il sito delle pp. 261–287. Series: IES Yearbook 2014. Allumiere di Monteleo e l’organizzazione della Bianchi, G. (2010) ʹDominare e gestire un territorio. produzione fra tardo Medievo e prima Età Modernaʹ, Ascesa e sviluppo delle ‘signorie forti’ nella in Arthur, P. and Imperiale, M.L. (eds) VII congresso toscana del Centro Nord tra X e metà XII nazionale di archeologia Medievale (Vol. 1). Firenze: secoloʹ, Archeologia Medievale, XXXVII, pp. 93–103. All’Insegna del Giglio, pp. 395–400. Boisseuil, D. and Chareille P. (2009) ʹL’exploitation del EPA (2007) SW–849 Test Method 6200: field portable x-ray l’alum en Toscane au début du XVI siècle: l’alunière fluorescence spectrometry for the determination of de Monterotondo et la société de Rinaldo Tolomeiʹ, elemental concentrations in soil and sediment. Available Mélanges de l’école française de Rome. Moyen-Âge, 121 at: https://www.epa.gov/hw-sw846/sw-846-test- (1), pp. 9–28. method-6200-field-portable-x-ray-fluorescence- Dallai, L. (2014), ʹL’allumiera di Monteleo nel territorio spectrometry-determination (Accessed: 28 March di Monterotondo Marittimoʹ, Mélanges de l’école 2017). française de Rome. Moyen-Âge, 126 (1), pp. 245–258. Marasco, L. (2013) ‘La Castellina di e le Dallai, L., Bardi, A., Briano, A., Buonincontri, M.P., fortificazioni di terra nelle pianure costiere della Buono, M., Fineschi, S., Poggi, G., Ponta, E., Rossi, M.,

145 M. Matsumoto and E. Uleberg (eds). CAA2016: Oceans of Data

Maremma settentrionale’, Archeologia Medievale, XL, Potts, P. J. and West, M. (eds) (2008) Portable X-ray pp. 57–68. fluorescence spectrometry: capabilities for in situ Oonk, S., Slomp, C.P., and Huisman, D.J. (2009) analysis. Cambridge: RCS Publishing. ʹGeochemistry as an aid in archaeological Shugar, A.N. and Mass, J.L. (eds) (2012) Handheld XRF for prospection and site interpretation: current issues art and archaeology. Leuven: Leuven University Press. and research directionsʹ, Archaeological Prospection, Series: Studies in Archaeological Sciences 3. 16 (1), pp. 35–51. doi: 10.1002/arp.344.

146